Progress OpenEdge AI Assistant

MCP stands for Model Context Protocol, an open standard that lets AI applications (like LLM-based agents) securely connect to external tools, data, and services in a consistent way.

An MCP connector is a software component that connects your IDE (e.g., Visual Studio Code, Windsurf, Cursor, etc.) or Command Line Interface (CLI)-based AI assistant to an MCP server and its resources. The connector translates natural language prompts into machine readable requests, transmits them to the MCP server, and returns results - such as generated code, refactored outputs or context-aware recommendations, and does so directly into your development environment.

The Progress OpenEdge AI Assistant is a product that links AI coding assistants to your specific OpenEdge ABL codebase and environment using the Model Context Protocol, so the AI can accurately generate, refactor, and modernize ABL code instead of guessing like a generic copilot. It acts as the “bridge” between your IDE (for example VS Code, Windsurf, Cursor) and an MCP server that has access to your ABL code, rules, and documentation, providing context‑aware, standards‑compliant assistance directly inside the tools you already use.

The OpenEdge AI Assistant went into general availability in January 2026.

The OpenEdge AI Assistant is offered as two separate deliverables that can be accessed from:

• Progress ESD (Electronic Software Download) files
• Progress hosted solution on PDC (Progress Data Cloud) portal

Follow these steps to access the Connector from ESD:

1. Login to ESD with your Progress credentials
2. Navigate to the OpenEdge 12.8.x page on ESD: https://downloads.progress.com/ProductLine/1/Product/1
3. Click the "AI Coding Assistance – ABL Syntax and Rules" link, to accept the EULA, and download the available files.

Follow these steps to access the OpenEdge AI Assistant from PDC:

1. Complete the "Request Access" form: https://www.progress.com/openedge/openedge-ai-assistant/contact
2. Complete the access request form and accept the End User License Agreement. The OpenEdge team will review your request and verify eligibility, typically within one business day.
   
3. Register on Progress Data Cloud to access the OpenEdge AI Assistant: https://docs.progress.com/bundle/openedge-ai-coding-assistant/page/Register-for-Progress-Data-Cloud-user-account.html

Note: You will receive an email to register on PDC once you have been approved to join the OpenEdge Customer Validation Program (CVP).  

Yes, the OpenEdge AI Assistant can analyze existing ABL code, follow organization‑specific standards, and assist with modernization initiatives – such as converting procedural code to OOABL, adding structured error handling, and improving design patterns.

The OpenEdge AI Assistant integrates seamlessly with multiple Integrated Development Environments (IDEs), including VS Code, Windsurf, Antigravity, Amazon Q, Cursor, Continue, and more – enabling AI-powered code assistance and rule-based enhancements across environments.

The OpenEdge AI Assistant does not save, store, or share any customer code, data, or intellectual property. The connector itself only provides an information repository containing OpenEdge ABL syntax, language rules, and reference material.

Any data a user chooses to provide to or receive from their GenAI coding assistant (such as Windsurf, Cursor, GitHub Copilot, or similar tools) is governed entirely by that GenAI provider’s terms, policies, and data handling practices — not by Progress.

For partners or customers with strict data residency or IP protection requirements where no code may leave their environment under any circumstance, Progress strongly recommends not using the cloud hosted OpenEdge AI Assistant. Instead, these customers should use the ESD downloaded ABL syntax and rules (.md) files directly with their chosen GenAI provider. This approach delivers very similar AI assisted development results while ensuring that all code and data remain fully within the customer’s-controlled environment.

For complete details regarding permitted use, deployment models, and data handling responsibilities, please review the OpenEdge AI Assistant End User License Agreement (EULA). See attachment.

Communication between the IDE or AI coding assistant and the MCP Connector’s MCP endpoint uses HTTPS/TLS (encryption in transit).

TLS is the default and recommended configuration when you register the MCP server in the IDE and should be mandatory for production use.

No, the OpenEdge AI Assistant is not designed to save/store/retain customer‑provided prompts or metadata at rest. To the extent your IDE or local environment logs activity, any such files are under your control and can be protected with your standard disk and backup encryption; there is no separate OpenEdge AI Assistant prompt store that Progress manages.

Customer input is handled transiently by the OpenEdge AI Assistant as part of interactive IDE sessions and is not retained by the connector beyond that processing. Any longer‑term history is governed by the IDE (e.g., chat window history) or by your chosen GenAI provider, not by the OpenEdge AI Assistant, and there is no product‑specific retention schedule for prompts within the OpenEdge AI Assistant.

The OpenEdge AI Assistant does not itself host or train an LLM and does not independently persist or process your code or IP for model training. Any retention, training, or secondary use of your prompts and code is governed solely by your selected GenAI provider and IDE configuration; the connector simply brokers requests between the IDE, MCP Connector (server) and those back‑end services.